Return to Website

Number Watch Web Forum

This forum is about wrong numbers in science, politics and the media. It respects good science and good English.

Number Watch Web Forum
Start a New Topic 

What follows is what I tried to post earlier.
This caused me to raise the question about infiltration of spam programs by the eco-activists and is now hopefully resolved by the changes made to the spam filter by our bending author.

NUmberwatch spam

An article in yesterday's (now some time back but accessible on the DT website under a different headline to that with which it was published)DT caught my eye and had me going for a while: "One small step for man, one giant blow for conspiracies".

This is the story of how a "Scientist" has developed a mathematical analysis that relates the length of time before frauds are exposed based on the number of people in the know.

Now in general I'd be inclined to agree that the more people who share a secret the shorter the time it will remain a secret.
In general.
But wartime produced some remarkable examples of well kept secrets with an inordinate number of people in the know. And some secrets last about five minutes with only two people in the know.

So I'm curious as to how this can be reliably mathematically quantified.
None the less, there is now a calculation and this has been applied to a few examples.

First up is the Moon landing.
With 411,000 employees the formula apparently determines that if it was a fake it would have been exposed within 3 years and 8 months. (wasn't it? when did the first claims it was faked arise?)

411,000 employees sounds like the popular big number but how many were in a party to the secret? Is there some weighting to discriminate between mission control staff, project directors and the tea ladies, parking attendants and tour guides?

And how do you test the calculation? I would have thought you needed some examples where the truth is known beyond any shadow of doubt to the researcher. And a lot of examples. (Added comment today: and sadly it appears there were only 3 as the debunking article Dave Gardner found shows).

But I gave up worrying about this when I got to the bit about +GW. his leads to the supposition that where most rational people may be thought to view the event as true, and the fake theory ludicrously false, by the time we get to A*W we are expected to say "Oh! *GW is true because of the number of people involved and here is a calculation that proves it."
(the "*" are intended to test if the "Spam" determination that has blocked this post seven times already and in different forms is keyed on any particular terms).

That's about when I went on to the crosswords and gave up on questioning the validity of the research or whether or not this was good use of the good doctor's time and grant money, when the suspicion came that the whole exercise was intended as yet another way to tackle the deniers.

PS I am not the only one to pick on these particular weaknesses. And there are many more including, it appears, some basic statistical errors. If I were a statistician I would campaign to have any publication that uses statistics vetted by a statistician prior to publication.

Re: Conspiracies.....

I would agree that the probable motivation behind the Grimes PLOS One paper, "On the Viability of Conspiratorial Beliefs", is to discredit AGW sceptics.

Grimes paper

Stephan Lewandowsky, now a professor at Bristol University, is the leading advocate of the idea that AGW sceptics are just conspiracy theorists and are liable to believe in a range of other conspiracy theories. (The origin of this particular idea probably goes back to Al Gore in his "Inconvenient Truth" film). The Grimes paper references four Lewandowsky papers, thanks Lewandowsky in the 'Acknowledgements' section, and the first two words in the abstract of the paper are "conspiratorial ideation", very similar to the obscure phrase "conspiracist ideation" which appears in the title of two of Lewandowsky's papers. It is a reasonable assumption that Grimes shares Lewandowsky's views.

In fairness to grant agencies, I don't think they actually funded Grimes for this paper. In the paper it states "Funding: The author has no support or funding to report". Also to publish in PLOS One, an author is charged a publication fee of $1495, which presumably Grimes has paid himself (PLOS One is one of a small number of scientific journals that cover their costs by getting the author of a paper to pay rather than the reader or a library).

It depends on how you define 'conspiracy' as to whether AGW can be regarded as a conspiracy. If we adopt the definition used in Grimes paper:
"Conspiratorial ideation is the tendency of individuals to believe that events and power relations are secretly manipulated by certain clandestine groups and organisations", that secrecy element isn't really present for climate science. Climate science is actually pretty transparent - for example they make predictions of future global temperatures that are always well above what actually happens, suggesting their computer models are not valid to anybody who cares to notice. I would regard AGW as being more of a charade than a conspiracy.

AGW reminds me a quite a bit of the situation with Jimmy Savile in the UK. There was no conspiracy in what Savile was doing decades ago - he was hiding in plain sight. Savile relied on the fact that people in management have surprisingly little observational skill and are often unwilling to go against what they think is the prevailing opinion. In this analogy, AGW sceptics would be the equivalent of the numerous people not in management who pointed out Savile was up to no good, and the British political class would be the equivalent of the BBC managers and hospital managers who didn't appear to notice anything or chose to overlook it.

Re: Conspiracies.....

Besides, Grimes assumptions are fatuous in the first place. So many conspiracies when later examined upon revelation, prove to have had 'whistle blowers' who could not get anyone on the outside to take any notice.